Reducing interference noise in a noisy speech recording has been a challenging task for many years yet has a variety of applications, for example, in handsfree mobile communications, in speech recognition, and in hearing aids. Traditional single-channel noise reduction schemes, such as Wiener filtering, do not work satisfactorily in the presence of non-stationary background noise. Alternatively, supervised approaches, where the noise type is known in advance, lead to higher-quality enhanced speech signals. This dissertation proposes supervised and unsupervised single-channel noise reduction algorithms. We consider two classes of methods for this purpose: approaches based on nonnegative matrix factorization (NMF) and methods based on hidden Markov models (HMM).  The contributions of this dissertation can be divided into three main (overlapping) parts. First, we propose NMF-based enhancement approaches that use temporal dependencies of the speech signals. In a standard NMF, the important temporal correlations between consecutive short-time frames are ignored. We propose both continuous and discrete state-space nonnegative dynamical models. These approaches are used to describe the dynamics of the NMF coefficients or activations. We derive optimal minimum mean squared error (MMSE) or linear MMSE estimates of the speech signal using the probabilistic formulations of NMF. Our experiments show that using temporal dynamics in the NMF-based denoising systems improves the performance greatly. Additionally, this dissertation proposes an approach to learn the noise basis matrix online from the noisy observations. This relaxes the assumption of an a-priori specified noise type and enables us to use the NMF-based denoising method in an unsupervised manner. Our experiments show that the proposed approach with online noise basis learning considerably outperforms state-of-the-art methods in different noise conditions.  Second, this thesis proposes two methods for NMF-based separation of sources with similar dictionaries. We suggest a nonnegative HMM (NHMM) for babble noise that is derived from a speech HMM. In this approach, speech and babble signals share the same basis vectors, whereas the activation of the basis vectors are different for the two signals over time. We derive an MMSE estimator for the clean speech signal using the proposed NHMM. The objective evaluations and performed subjective listening test show that the proposed babble model and the final noise reduction algorithm outperform the conventional methods noticeably. Moreover, the dissertation proposes another solution to separate a desired source from a mixture with arbitrarily low artifacts.  Third, an HMM-based algorithm to enhance the speech spectra using super-Gaussian priors is proposed. Our experiments show that speech discrete Fourier transform (DFT) coefficients have super-Gaussian rather than Gaussian distributions even if we limit the speech data to come from a specific phoneme. We derive a new MMSE estimator for the speech spectra that uses super-Gaussian priors. The results of our evaluations using the developed noise reduction algorithm support the super-Gaussianity hypothesis. / <p>QC 20130916</p>

This thesis identifies a trend in recent postmodern dystopian fiction for writers to metafictionally dwell on the place of literature in a future context. This trend springs from similar concerns present in the two most influential dystopian novels of the 20th century, Brave New World and Nineteen Eighty-Four. Yet, unlike Huxley and Orwell, for whom the marginalization of literature is merely one symptom of the hegemonic control oppressing these future societies, the postmodern writers I identify situate the book’s future disappearance at the epicenter of culture’s demise. In Gary Shteyngart’s Super Sad True Love Story (2010), electronic technologies have virtually eradicated print literature and the novel’s protagonist, Lenny, mourns the changes in social interactions he sees this shift in technology bringing about. In Jeanette Winterson’s The Stone Gods (2007), marginalized book-lovers see the devastation humanity continuously wreaks on the environment as a product of culture’s disdain for literature.

Depuis quelques années déjà, les films de savon représentent un cadre idéal pour étudier les écoulements 2D et également la turbulence bidimensionnelle ( Couder). Il s'agit ici d'étudier la convection thermique dans une demi-bulle de savon. Le gradient thermique entre l'équateur et le pôle engendre une turbulence autour de l'équateur et donne naissance à des structures tourbillonnaires uniques de grandes tailles proches du pôle. Je mettrais en évidence une dynamique identique entre les tourbillons formés à la surface de la demi-bulle et les systèmes que l'on observe dans la nature ( les ouragans), à savoir un comportement super-diffusif. Une étude de la turbulence engendrée par la convection thermique a également été réalisée. La température présente une transition étonnante d'un comportement intermittent pour les gradients de température faible, à un comportement non intermittent lorsque l'on accroît le gradient de température. Cette dynamique est corrélée avec celle du champ de vitesse qui montre également cette transition.

Cette thèse concerne l'étude de la dynamique d'impacts de gouttes, dans des situations de friction réduite entre le substrat solide et la goutte liquide. Cette diminution de friction s'est faite au moyen d'un film d'air inséré entre le liquide et le solide. Il existe plusieurs stratégies permettant l'existence de ce film d'air : la première est d'utiliser le phénomène de caléfaction, ou effet Leidenfrost : un liquide approché d'une surface chauffée au delà d'une température critique s'évapore suffisamment rapidement pour pouvoir léviter sur sa propre vapeur, et ainsi être isolé de la surface solide. Dans certaines conditions, les surfaces super-hydrophobes micro-texturées permettent au liquide de rester dans un état "fakir", c'est à dire de n'être en contact qu'avec le sommet de micro-piliers, le reste du liquide demeurant au dessus d'un coussin d'air. Enfin, il a également été constaté que l'écoulement d'air engendré par le mouvement d'une surface solide peut induire une force de portance sur une goutte, et ainsi lui permettre de léviter au dessus de cette surface

Le projet SACER vise à répondre aux demandes d'Airbus qui ont besoin de disposer de données décrivant le comportement d'un avion ou d'un satellite avant commercialisation ou lancement. Pour mieux répondre à cette demande, un réseau de capteurs sans fil remplacerait les équipements de test filaires existants. Le but est d'apporter des avantages tels qu'une réduction de poids, de coût et de connectique. Pour notre part, nous n'avons travaillé que sur l'application aéronautique. Pour alimenter les capteurs autonomes sans fil embarqués, dans le cadre de cette thèse, il faut concevoir une architecture de récupération d'énergie, un système de stockage de l'énergie ainsi que un circuit de gestion de ces énergies. La principale contrainte pour le système est qu'il doit pouvoir fonctionner de -50C à 100°C, tout en délivrant une puissance de sortie de 3 watts. De plus, l'épaisseur du système doit être inférieure à 3,2 mm. Pour notre travail, nous avons cherché, dans un premier temps, la meilleure solution possible sur le choix du type de cellules solaires. Le résultat sur les tests des cellules à différentes températures et irradiations dans les conditions de notre application est présenté. Dans un second temps, nous avons testé plusieurs types de systèmes de stockage d'énergie aux températures extrêmes. Enfin, la conception de l'architecture pour la gestion de l'énergie (vue d'ensemble des panneaux photovoltaïques, d'un circuit MPPT, des super condensateurs, et d'un régulateur) est présentée.

Une méthode de calcul non-linéaire d'écoulement autour de structures minces par la méthode des singularités est présentée. Le concept de minimisation est développé et appliqué à des configurations sub-cavitantes, partiellement cavitantes et super-cavitantes. Ce concept est relatif aux erreurs numériques commises sur les débits de fuite. Le calcul non-linéaire de la géométrie de la poche de cavitation est obtenu par une procédure itérative. La méthode est validée par différentes comparaisons des résultats avec les résultats de référence issus de la littérature. Le cas de la grille d'aubes a été traité, les résultats comparés à ceux obtenus avec une méthode d'éléments finis. La comparaison est excellente. Le code de calcul a été étendu au calcul 3D de l'écoulement autour d'une aile partiellement cavitante et super-cavitante en présence d'une surface libre ainsi qu'au calcul 3D d'une hélice sub-cavitante, suivie de son sillage, en écoulement tournant. Les résultats sont confrontés aux résultats expérimentaux.

Robust speech recognition in acoustic environments that contain multiple speech sources and/or complex non-stationary noise is a difficult problem, but one of great practical interest. The formalism of probabilistic graphical models constitutes a relatively new and very powerful tool for better understanding and extending existing models, learning, and inference algorithms; and a bedrock for the creative, quasi-systematic development of new ones. In this thesis a collection of new graphical models and inference algorithms for robust speech recognition are presented. The problem of speech separation using multiple microphones is first treated. A family of variational algorithms for tractably combining multiple acoustic models of speech with observed sensor likelihoods is presented. The algorithms recover high quality estimates of the speech sources even when there are more sources than microphones, and have improved upon the state-of-the-art in terms of SNR gain by over 10 dB. Next the problem of background compensation in non-stationary acoustic environments is treated. A new dynamic noise adaptation (DNA) algorithm for robust noise compensation is presented, and shown to outperform several existing state-of-the-art front-end denoising systems on the new DNA + Aurora II and Aurora II-M extensions of the Aurora II task. Finally, the problem of speech recognition in speech using a single microphone is treated. The Iroquois system for multi-talker speech separation and recognition is presented. The system won the 2006 Pascal International Speech Separation Challenge, and amazingly, achieved super-human recognition performance on a majority of test cases in the task. The result marks a significant first in automatic speech recognition, and a milestone in computing.

Providing an energy efficient recycle for the Teflon® synthesis process is of great interest due to environmental and economic reasons. This recycle step involves separating CO2 from a stream containing scCO2 and valuable monomer (C2F4). Membranes provide economical and environmental friendly separations compared to conventional methods (e.g. distillation, amine absorption). Therefore, I am investigating membrane materials that are well-suited for this important separation. Developing a robust membrane that can withstand the aggressive scCO2 environment (~1070 psi of CO2) is a key challenge. Supercritical CO2 swells traditional polymeric membrane materials, thereby increasing segmental mobility of the polymer chains which leads to a decrease in separation capacity. There have been no polymeric membrane materials identified in the literature which are suitable for this separation. In this work, I have identified an advanced polymer, Torlon® (a polyamide-imide), that solves this problem. After determining the appropriate material, it is important to choose a membrane morphology that is industrially desirable. The asymmetric hollow fiber membrane morphology provides the highest productivity. I have successfully produced defect-free asymmetric hollow fiber membranes using Torlon® that withstand high pressure feeds. These membranes have been shown to provide selective separations under scCO2 conditions without being plasticized. To further improve the separation performance of Torlon® membranes, mixed matrix concept was explored. Zeolite 4A, which is relatively more permeable and selective compared to Torlon®, was chosen as the sieve material. Mixed matrix membranes from Torlon® and zeolite 4A were made and their separation performance was measured. Based on these experimental measurements and Maxwell modeling, challenges in making successful mixed matrix membranes were identified and feasible solutions for these challenges are suggested.

Les récentes avancées en microscopie de fluorescence (augmentation de résolution spatiale, correction des aberrations induites par l'échantillon...) ouvrent de nombreuses perspectives en imagerie biologique. Mais pour que ces techniques se répandent, il est nécessaire de réduire les contraintes qu'elles imposent sur la préparation des échantillons, sur les sondes fluorescentes, sur la complexité de mise en oeuvre ou la cadence d'acquisition. Pendant ma thèse, j'ai travaillé sur le développement et l'amélioration de quelques-unes de ces techniques. Dans un premier temps, deux systèmes d'illumination structurée ont été réalisés dans le but d'accélérer la cadence de réalisation de coupes optiques pour l'observation d'échantillons épais vivants. Une deuxième partie de mon travail a concerné la mise en place d'une nouvelle approche de reconstruction en illumination structurée afin de dépasser la limite de résolution imposée par le microscope. Cet outil permet de réduire le nombre d'images requises pour la reconstruction d'une image super-résolue, ce qui ouvre la voie à une augmentation de la cadence de réalisation de ces images. Un montage d'optique adaptative a également été développé afin de corriger les aberrations introduites par les échantillons épais, permettant une amélioration des capacités d'imagerie en profondeur. L'accent a été mis sur la protection de l'échantillon, grâce à l'utilisation de billes fluorescentes pour déterminer la correction à appliquer sans illuminer l'échantillon. Enfin, un instrument de mesure du rendement quantique de fluorescence dédié à la caractérisation de fluorophores infrarouges a été mis en oeuvre afin de pallier les limitations des méthodes de mesure classiques dans le proche infrarouge.

Subject identification from surveillance imagery has become an important task for forensic investigation. Good quality images of the subjects are essential for the surveillance footage to be useful. However, surveillance videos are of low resolution due to data storage requirements. In addition, subjects typically occupy a small portion of a camera's field of view. Faces, which are of primary interest, occupy an even smaller array of pixels. For reliable face recognition from surveillance video, there is a need to generate higher resolution images of the subject's face from low-resolution video. Super-resolution image reconstruction is a signal processing based approach that aims to reconstruct a high-resolution image by combining a number of low-resolution images. The low-resolution images that differ by a sub-pixel shift contain complementary information as they are different "snapshots" of the same scene. Once geometrically registered onto a common high-resolution grid, they can be merged into a single image with higher resolution. As super-resolution is a computationally intensive process, traditional reconstruction-based super-resolution methods simplify the problem by restricting the correspondence between low-resolution frames to global motion such as translational and affine transformation. Surveillance footage however, consists of independently moving non-rigid objects such as faces. Applying global registration methods result in registration errors that lead to artefacts that adversely affect recognition. The human face also presents additional problems such as selfocclusion and reflectance variation that even local registration methods find difficult to model. In this dissertation, a robust optical flow-based super-resolution technique was proposed to overcome these difficulties. Real surveillance footage and the Terrascope database were used to compare the reconstruction quality of the proposed method against interpolation and existing super-resolution algorithms. Results show that the proposed robust optical flow-based method consistently produced more accurate reconstructions. This dissertation also outlines a systematic investigation of how super-resolution affects automatic face recognition algorithms with an emphasis on comparing reconstruction- and learning-based super-resolution approaches. While reconstruction-based super-resolution approaches like the proposed method attempt to recover the aliased high frequency information, learning-based methods synthesise them instead. Learning-based methods are able to synthesise plausible high frequency detail at high magnification ratios but the appearance of the face may change to the extent that the person no longer looks like him/herself. Although super-resolution has been applied to facial imagery, very little has been reported elsewhere on measuring the performance changes from super-resolved images. Intuitively, super-resolution improves image fidelity, and hence should improve the ability to distinguish between faces and consequently automatic face recognition accuracy. This is the first study to comprehensively investigate the effect of super-resolution on face recognition. Since super-resolution is a computationally intensive process it is important to understand the benefits in relation to the trade-off in computations. A framework for testing face recognition algorithms with multi-resolution images was proposed, using the XM2VTS database as a sample implementation. Results show that super-resolution offers a small improvement over bilinear interpolation in recognition performance in the absence of noise and that super-resolution is more beneficial when the input images are noisy since noise is attenuated during the frame fusion process.